Shell & Tube Heat Exchangers - Fouling Factors & Design Margins

My thoughts at a Refresher Course on Heat Transfer organised by Indian Chemical Council yesterday.

1. Huge sums of money are spent in treating the water used for cooling tower make-up and also to control and monitor the quality of circulating cooling water. But despite successfully altering the scaling nature of the circulating cooling water, process engineers are reluctant to shirk off the legacy of the conservative fouling factors propounded in TEMA as far back as 1941.

2. On top of this 15-20% safety margins are added on flow and heat duty while designing heat exchangers. There are various reasons for this:

• Uncertainties in process
• Uncertainties in physical properties
• Providing for future expansion
• Fear of not meeting guarantees
• Mistaken notion that this will improve reliability

3. While revamping and de-bottlenecking plants, shell and tube heat exchangers are rarely ever replaced. I know a large petrochemical plant whose capacity was increased by 150% from 80,000 tpa to 120,000 tpa by changing catalyst, pump impellers, control valve trims and air cooler motors, but leaving all of the 50 odd shell and tube heat exchangers untouched. The hidden fat in heat exchanger design is obvious.

4. Using the real life cooling water chemistry in the newly introduced General Cooling Water Fouling model of HTRI reveals that the TEMA fouling factors are higher by 200-300 %. The accuracy provided by sophisticated and expensive models and software would be nullified if we keep sticking to the ultra-conservative fouling factors from TEMA.

5. Operating a perfectly optimised shell and tube heat exchanger at 80% of its design load would only accelerate fouling because of the lower velocities and hence thicker boundary layer film. Do we need to spend more money in over-designed shell and tube heat exchangers and then operate them under sub-optimal conditions?

6. With TEMA fouling factors in place, there is no need for further safety margin. 20% more flow can easily be pushed through a heat exchanger since the necessary hydraulic buffer is usually available in the system. For engineers who still like to be conservative, providing 10% margin on the calculated surface area is a better option than designing with 10% margin on the flow.